Based on my understanding, spzeros(m,n) would create a sparse matrix of zeros with size m*n. I checked and found that calling isempty with a matrix like this would return false. What I do not understand is that, based on this link, Sparse Arrays · The Julia Language,
spzeros(m,n) is empty? …
isempty is meant to tell you if a collection has any elements in it (docs). A sparse matrix is an array, which Julia sees as a collection too, just like a vector v = [0.0, 1.0] is both a 1-dimensional array and a collection of two things (namely 0.0 and 1.0). So spzeros(m,n) creates an n\times m sparse matrix which has n*m elements – and so it’s not empty-- the elements just all happen to be zero.
help?> isempty
search: isempty
isempty(collection) -> Bool
Determine whether a collection is empty (has no elements).
A spare matrix of all zeros has elements (all zeros) as long as its size is bigger than (0,0). This is also consistent with how normal arrays behave.
The function nnz may be what you want. For a sparse matrix A that does not store any nonzero element, there exists nnz(A) == 0